In reply to:

Gapless playback of properly-constructed MP3 files has been improved. Note
that this is dependent upon how the MP3s were created - most MP3 encoders
will add artificial silence to the beginning and end of each track as it
is being created. The car player will no longer insert its own gaps between
two consecutive tracks on playback.


This is great news! Now since it's a beta test I'll get to see if this is really implemented well. So here's what I will want answered, either by my own tests or from the people who might already know the answer.

Let's say I create a single, large MP3 file and split it into several pieces---for example, I encode an entire album and split the file into individual traks. Now suppose I play them in their original consecutive order on an empeg. Will it sound exactly as if the file was never split in the first place?

It is important to note that it is not sufficient to simply insure that the MP3 decoder does not insert any additional gaps between the MP3 files. An additional necessary condition is that the MP3 decoder's internal state is not reset between MP3 files, either. In other words, the computational section of the MP3 decoder must not even know where the file boundaries are. If the internal state of the MP3 decoder is reset between files, then a certain amount of information may be lost about the first few frames after a split.

As I have discussed in previous threads, this is the only practical way that I know of, for most if not all MP3 encoding engines, to insure true gapless playback: specifically, to encode an entire album as a single MP3 file, and then split the file along frame boundaries where the track boundaries are supposed to be.

Clearly, this is different from splitting the uncompressed .WAV file into pieces, and then encoding each one separately. One reason is somewhat obvious: if the length of a .WAV file is not an exact multiple of the MP3 frame size (1/75th of a second, I believe?) then silence will be inserted at the beginning or the end of the file to make it so. And yet, even careful, MP3-sensitive .WAV file splitting will not be enough: an MP3 encoder will often add additional silence to the beginning and/or the end of the file upon encoding despite this.

This is because each MP3 frame is not encoded independently of each other. So the sound produced during a particular MP3 frame depends upon the frames that surround it. Put another way, some of the information used to create the sound in a given MP3 frame is placed before or after that frame in the file. If this information would then run off the end of the file, blank space is inserted there to hold it.

So, the question is: will 2.00 pass this test---if not now, by the release? I am truly excited that it can!

Michael Grant
12GB Green
080000266
_________________________
Michael Grant 12GB Green 080000266